Real-time Monitoring System for Active Volcanoes in Costa Rica
Research Partners:
- Mauricio Mora (Red Sismológica Nacional – UCR)
- Javier Pacheco (Observatorio Vulcanológico y Sismológico de Costa Rica – UNA)
Lead Researcher at CeNAT:
Esteban Meneses
Description:
Costa Rica sits near a subduction zone of two tectonic plates. It is therefore not unusual to have a high seismic and volcanic activity in the country. The latter is reflected on the recent volcanic agitation of several volcanoes. For instance, Turrialba Volcano went through a cycle of eruptions that caused major commotion in the country: roads and airports had to be shut down, crops were lost, touristic trips were cancelled, and so on. Consequently, it is crucial to improve our understanding of active volcanoes to ameliorate the effects of volcanic activity. This project aims at implementing a real-time pipeline that transports seismic signals from sensor networks at volcanoes, analyzes and classifies those signals, and provides a profiles of each volcano along with potential warnings. The use of Machine Learning techniques and High Performance Computing platforms is essential in achieving our goal.
Infrastructure of Computational Plasma Simulation for Design and Verification of Magnetic Confinement Devices type Stellarator
Research Partners:
- Ricardo Solano Piedra (Laboratorio de Plasmas para Energía de Fusión y Aplicaciones – TEC)
- Alonso Araya Solano (Laboratorio de Plasmas para Energía de Fusión y Aplicaciones – TEC)
- Esteban Pérez (Laboratorio de Plasmas para Energía de Fusión y Aplicaciones – TEC)
Lead Researcher at CeNAT:
Diego Jiménez
Description:
This project seeks to fulfill the need for more complex simulation models and the ability to generate scientific visualizations by constructing a computational infrastructure that unifies the workflow of researchers. In doing so, the process of design and verification for new magnetic-confinement Stellarator devices in Costa Rica will be improved through advanced computing.
Deep MIML for Tropical Bird Species Classification
Research Partners:
- Roberto Vargas (UNED – LIIT)
- Danny Alfaro (UNED – LIIT)
Lead Researcher at CeNAT:
Jorge Castro
Description:
Deep learning techniques are proposed together with the Multiple Instance Multiple Label (MIML) framework to identify tropical bird species in acoustic field recordings. The main goal of the project is to develop a scalable solution to monitor birds reducing the amount of manual work typically needed to train a deep neural network.
Modelling Urban Mobility in Costa Rica with Waze
Research Partners:
- State of the Nation Program, San José (PEN)
- Ministry of Public Works and Transportations (MOPT)
Lead Researcher at CeNAT:
Mariana Cubero
Description:
Study of the urban mobility and traffic dynamics in Costa Rica using as a main data source reports of traffic conditions in Waze. We model and characterize traffic conditions to understand the dynamics of mobility and the challenges faced by the population.
ITS Data Analysis from Endophytic Fungi of Coffee Plants using Specialized Databases
Research Partners:
- Efraín Escudero Leiva (CIPRONA & CENIBiot)
- Prisilla Chaverri (CIPRONA & Department of Plant Science and Landscape Architecture, University of Maryland)
Lead Researcher at CeNAT:
Maripaz Montero
Description:
Design and standardization of a pipeline for the analysis of ITS sequences of endophytic fungi, which covers from basic sequence processing to phylogenetic analysis.